A randomised, prospective, placebo-controlled phase III multicentre clinical trial (KyberSept) has been performed to test the efficacy of high-dose
antithrombin therapy in patients with
severe sepsis. Concomitant low-dose
heparin has been routinely given in two thirds of patients for
deep vein thrombosis prophylaxis. This study analyses
heparin -
antithrombin interactions in terms of long-term mortality, adverse events, and thromboembolic events. From a total of 2,314 patients with
severe sepsis (placebo: n = 1,157;
antithrombin: n = 1,157) 1,616 patients (placebo: 811,
antithrombin: 805) received
heparin concomitantly with study
drug (
antithrombin 30,000 IU) over four days, whereas 698 patients (346 and 352, respectively) did not. In patients with no concomitant
heparin, 28-day mortality was lower with
antithrombin than with placebo (37.8% vs. 43.6%; absolute reduction: 5.8%; risk ratio: 0.860 [0.725-1.019]), which increased until day-90 (44.9% vs. 52.5%; absolute reduction: 7.6%; risk ratio: 0.851 [0.735-0.987]). In patients with concomitant
heparin, no effect of
antithrombin on mortality was seen (28-day mortality: 39.4% vs. 36.6%; absolute increase: 2.8%; risk ratio: 1.08 [0.96-1.22]). Frequency of use of concomitant
heparin increased during conduct of the study. Increased
bleeding incidences were reported with
antithrombin plus concomitant
heparin as compared to
antithrombin alone. Rates of thromboembolic events were similar when
antithrombin was given with or without concomitant
heparin. In the treatment of
severe sepsis, high-dose
antithrombin may sufficiently protect against development of
venous thromboembolism when no concomitant
heparin is given. Combined administration of the two increases
bleeding risk and probably abolishes efficacy of
antithrombin.